list_sleepwalker_testsLABSCO SUMMARY
The two halves connect: it shows what the engines say, then scores the pages that decide it.
A visibility check on its own tells you that Perplexity recommends a competitor. Pairing it with content scoring on your own pages turns that into something you can act on. Results persist in one workspace, so a run an agent starts is reviewable in the app later or queryable from code — which matters when the person acting on the finding is not the one who ran it. Two practical notes: runs are queued, so expect to check status rather than wait, and the actions that cost credits are separate from the ones that just read what you already paid for.
What it is
A hosted service with two connected workflows — AI visibility checks across the major answer engines, and content intelligence over your public pages — reachable from an MCP client, the app, an API or a CLI.
What you get
- create_sleepwalker_visibility_run — prompts run across ChatGPT, Perplexity, Grok and Gemini, capturing full answers, citations, competitors and mention types
- suggest_sleepwalker_visibility_prompts — the prompts worth testing, proposed rather than invented by you
- list_sleepwalker_visibility_models and list_sleepwalker_visibility_runs — which engines are covered, and what you have already run
- get_sleepwalker_visibility_run_status and cancel_sleepwalker_visibility_run, since runs are queued rather than instant
- get_sleepwalker_prompt_response — the actual answer an engine gave, not just a score
- serialize_sleepwalker_page_content — a public page turned into the content and technical snapshot an AI crawler would see
- score_sleepwalker_content and discover_sleepwalker_content_trends — depth and freshness scored, and demand identified
- get_sleepwalker_reports_by_url, get_sleepwalker_summaries and get_sleepwalker_consistent_recommendations — past work retrieved and the recurring advice pulled out
- list_sleepwalker_tests and get_sleepwalker_test_results, for saved checks you re-run
Requirements
A Sleepwalker account. The hosted endpoint is https://mcp.sleepwalker.ai/mcp — connect over OAuth if your client supports it, or create a bearer token in the app for a custom client. Billable actions draw on prepaid Sleepwalker credits, so runs cost money while reading past results does not. The open-source CLI export command runs locally with no key at all.
Setup effort
Add the hosted endpoint, then sign in — paste it into your client, then approve the OAuth sign-in
USE CASES
01
Finding out whether AI search recommends you at all
Run the prompts a buyer would actually type across four engines and read the answers, citations and competitors named.
02
Working out which prompts are worth testing
Get suggestions for your brand rather than guessing at the phrasing yourself.
03
Seeing your page the way a crawler does
Serialize the URL and get both the content and the technical snapshot, in one call.
04
Deciding which page to rewrite first
Score depth and freshness across the candidates, then read the recommendations that keep recurring.
05
Tracking whether a change moved anything
Re-run a saved test and compare against the earlier report for the same URL.
06
Checking on a run that is still queued
Ask for its status, and cancel it if the parameters were wrong, rather than waiting it out.
18 tools
Listed in project docs. Your client may show a different list.
get_sleepwalker_test_resultsget_sleepwalker_reports_by_urllist_sleepwalker_visibility_runsget_sleepwalker_visibility_run_statuslist_sleepwalker_visibility_modelslist_sleepwalker_content_runsget_sleepwalker_content_run_statusget_sleepwalker_prompt_responseget_sleepwalker_summariesget_sleepwalker_consistent_recommendationsserialize_sleepwalker_page_contentsuggest_sleepwalker_visibility_promptscreate_sleepwalker_visibility_runcancel_sleepwalker_visibility_rundiscover_sleepwalker_content_trendsscore_sleepwalker_contentcreate_sleepwalker_content_runBEFORE YOU INSTALL
What to know first
Tools you get
18 tools named in the project's documentation.
Requires
A client that can reach a remote MCP server over HTTP
Where it runs
Both documented clients connect to a hosted endpoint.
INSTALLATION
Connect your client
Server setup
Claude Desktop — OAuth
ENDPOINT URL
https://mcp.sleepwalker.ai/mcpAdd this hosted endpoint to your MCP client using the credentials listed under Requirements.
MCP Clients
Any MCP client — bearer token
ANY MCP CLIENT — BEARER TOKEN · JSON
{
"mcpServers": {
"sleepwalker": {
"command": "npx",
"args": [
"-y",
"mcp-remote",
"https://mcp.sleepwalker.ai/mcp",
"--header",
"Authorization:${SLEEPWALKER_MCP_AUTH_HEADER}"
],
"env": {
"SLEEPWALKER_MCP_AUTH_HEADER": "Bearer <YOUR_TOKEN>"
}
}
}
}Create an MCP bearer token in the Sleepwalker app ; that is <YOUR_TOKEN>. Keep the token private.
Values written as ${…} are filled in only by clients that expand variables.
